Our immune systems are walking a tightrope. Every day the cells that comprise our immune system are making decisions that protect us from infectious diseases. Regrettably, however, this same system has the potential to cause disease. Problems arise when we attack our own tissues leading to autoimmune conditions, or we respond too strongly to innocuous threats, as in allergy to peanut, or pollens. How do cells of the immune system cooperate and make the correct decisions that protect us from infection while avoiding responding to our own tissues? Our lab is intent on solving this general problem.
We use careful measurement of immune cell responses under highly controlled conditions to discover how they behave with mathematical precision. We use this knowledge to inform development of computer-based predictive models of immunity. We use these models in turn to explore the decision making potential of the immune system and develop new treatments for a broad range of diseases that include lymphoma, immunodeficiencies, autoimmunity and allergy.
At its core our mission is to help formulate a new understanding of immunity using methods borrowed from physics and mathematics. We aim to develop a quantitative version of the immune system, effectively updating the amazing, but 60-year-old, clonal selection theory created by former WEHI director and Nobel laureate, Sir Macfarlane Burnet. Completing this mission will greatly assist the goal of identifying the causes of immune dysfunction and finding ways to diagnose and treat such problems.
Our lab has contributed a series of discoveries that advance our understanding of how the cells of the immune system are controlled and ‘add’ signals together to affect their fate.
In a series of studies over two decades we have shown how both T and B lymphocytes operate a ‘cellular calculus’ that takes input from receptor delivered signals to regulate their times to divide, die and differentiate.
We championed the view that cells behave as if made of modular mechanical components connected and altered by these signals. Such internal operators include division and death timers as well as differentiation fates controlled by both timers and division counters. Combining such operators into multiscale cell models help guide the experimental dissection of this otherwise very complex system.
Importantly, we have also discovered that cells themselves differ in their construction as a result of stochastic processes. This feature ensures populations of similar cells give rise to heterogeneous outcomes. Thus, our insights into how to model and predict immune cell responses take advantage of the modular construction of cells, their quantitative ability to process signals according to rules definable by a cellular calculus, and our discovery that the remarkable differences between similar cells can be interpreted within probabilistic principles.
Ultimately this work is aimed at completing a consistent theory for immunity that can be viewed as the modernization of the clonal selection theory.
Recent highlights include identification of how oncoprotein Myc serves as a heritable timer for controlling how long cells divide for after stimulation (See Heinzel et al. Nature Immunology 2017), and the identification of intersecting stochastic processes that regulate the choice of antibody isotype adopted by stimulated B cells (Horton et al. Immunity 2022).
The development of mathematical models of the T and B cell adaptive immune response has developed rapidly over the last several years and the probabilistic principles for codifying modules of cellular behavior have proved increasingly successful. All members of the lab work either with, or on improving such models directly. New resources are being developed as software for the immunology community. Experiments to inform the models and to test predictions are made from single cell tracking, from cell division and differentiation tracking and from fate mapping performed both in vitro and in vivo.
Team members: Evan Thomas, Ken Duffy (National University of Ireland).
There is still much to learn about the normal biology of both T and B lymphocytes in the immune response. We are particularly keen to understand how cells add signals together and adjust to changing levels and combinations of the many different cytokines and costimuli on offer during an immune challenge. To reveal this ‘cellular calculus’ we measure cell behavior, at single cell and population level to learn the rules of addition and to predict the net outcomes. The effects of drugs and genetic manipulations on modular components of the cell are also being measured with the aim to develop a principle for predictive immunotherapy and in silico drug screening.
The model team – Su Heinzel, Michelle Ruhle, Melissa Biemond, Evan Thomas
Our principles of cellular calculus and single cell behavior have been developed using model systems. We are asking whether the same principles operate for people and the quantitative methods we have developed could help stratify and screen for genetic deficiencies and susceptibilities. Importantly we aim to identify how multiple small quantitative changes in cellular circuits can add up to powerful immune disorders such as autoimmunity. Our first target is to examine Common Variable Immunodeficiency and move from there to more complex immune disorders.
Team members: Vanessa Bryant, Su Heinzel and Charlotte Slade
Our lab adopts ‘big picture’ goals aimed at understanding immunity at molecular, cellular and whole system level. Hence we typically work at the interface of experiment and theory.
We have a strong collaboration with probability expert Professor Ken Duffy, National University of Ireland. We also have greatly benefited from the input and ideas brought from physics and math-trained lab members.
We believe that the complex challenge of the immune system presents many analogous philosophical problems to those confronted by physics research in the past. We have been particularly inspired to reconcile probabilistic and deterministic theories of causation for cells. We find immune cell fate is highly deterministic once in train, but that at key transition periods stochastic processes are called upon to ensure multiple outcomes are pursued by the responding cells.